EEVblog #605 - Microphone Patterns

[EEVblog]

Doug Ford, former head designer from Rode Microphones continues with Part 2 of the microphone technology series by explaining the construction of noise cancelling Figure 8, cardioid, and hyper cardioid microphones. Also, how the polar patterns and responses relate to the physical construction, and how the frequency response is affected.
Proximity boost effect,
Microphone calibration is also discussed using a home made artificial voice speaker box.
And practical considerations about foldback wedges used in stage performances and how to avoid feedback using proper microphone technique.
And did Jimmy Barnes offer Doug a swig from his bottle of Vodka?

[dentaku]

I don't think I've ever heard of Jimmy Barnes before. I "Googled it with Bing" (to quote Scott Hanselman) to find out.

Anyway... This explains so much about mics that I've heard about many times but never really looked into.

[DrGeoff]

A significant use of fig-8 microphones in recording applications is in the MS (mid-side) microphone array.

This is particularly good for recording larger ensembles, where the room acoustice is desirable in the recording. In this case the fig-8 is positioned with the null facing the ensemble, and the lobes effectively pointing towards the room sides, and a second mono mic (either a cardioid or omni pattern) is positioned as close to the fig 8 as possible, aimed directly at the ensemble.

From this there are two mono signals (one fig-8 and one omni) which are recorded on two separate tracks. At mix time, a true stereo image (with variable width) can be created by taking the fig8 signal, duplicating it in another channel and inverting its polarity, and panning one hard left and the other hard right. Now increase the omni signal level to balance the image. The more omni mic signal that is introduced, the narrower the stereo image (ie more of the mono centre signal).

I was testing out a new field recorder (Zoom H6) using its included MS microphone on the Sydney Wind Symphony at a concert last year and here you can hear the effect of an MS recording and the accurate position of each of the instruments in the stereo field.
https://soundcloud.com/dr-geoff/el-camino-real

[Alexei.Polkhanov]

Hmm, I listened using my headphones and I felt like I am little deaf in my left ear - expected effect?

[DrGeoff]

Hmm, I listened using my headphones and I felt like I am little deaf in my left ear - expected effect?

Nope, the sound is occurring as per the ensemble. Some parts are on the left (french horns, flutes) and some are on the right (clarinets, low brass). Some are spaced across the back (percussion) and others are fairly central (oboes, bassoons, sax, trumpets, bass clari etc). The stereo field is balanced, however, so it should not be leaning one way or the other for any lengthy duration.

BTW, it was recorded just behind the conductor, so the field is fairly wide, and this piece does feature quite a bit of clarinet work from the front row so in parts it will sound more active on the right side, since that is where the players are, however this is augmented in parts by the flutes and french horns on the left (example around 7:55).

[Kompost]

I was testing out a new field recorder (Zoom H6) using its included MS microphone on the Sydney Wind Symphony at a concert last year and here you can hear the effect of an MS recording and the accurate position of each of the instruments in the stereo field.

The performance of this recorder is simply amazing. It seems to me that the highs are way softer than on H4, did you eq the track or do the H6 mics sound that way?

[Richard Crowley]

@DrGeoff, nice recording!  Great performance and handy little recorder. I will probably add one to my kit.

Mid-Side (MS) mics require a cardioid "mid" polar pattern. Note that using an omni as a "mid" element results in a cardioid facing hard left (or right, depending on your maths). This was shown by Mr. Ford in the latest video.

Using a cardioid "mid" polar, you can derive the Left and Right signals using the formulae:

Mid + Side = Left
Mid - Side = Right

Assuming the "positive" side of the figure-8 is facing left

It is interesting to note, however, that in most modern lower-cost (and even some insanely expensive) "MS" mics, the "MS" function is actually implemented with three cardioid mic elements. One facing left, one facing "mid", and the last facing right.  By adding L+C, you get a left polar pattern rather half-way between the front-facing capsule and the left facing capsule, and likewise on the right side.

The reason for using three cardioids vs. one cardioid and one figure-8 is that there are few decent studio-quality true figure-8 mic capsules with which to build microphones. I don't believe that Rode makes one, for example.

It would be interesting to discover how much difference you can hear between a LMR-style "MS" microphone and an ordinary "X-Y" coincident pair.

[DrGeoff]

@DrGeoff, nice recording!  Great performance and handy little recorder. I will probably add one to my kit.

Thanks Richard. The interchangeable mic heads and the 4 extra channels with XLR/TRS and phantom power make it a very useful piece of field kit. The

Mid-Side (MS) mics require a cardioid "mid" polar pattern. Note that using an omni as a "mid" element results in a cardioid facing hard left (or right, depending on your maths). This was shown by Mr. Ford in the latest video.

The MS mid mic can be either omni or cardioid. There are several articles about this. It depends on your application as to whether you want a rear rejection in the mono element (cardioid) or you want sound from the rear in the mono element (omni). The maths for the sides is the same. There are advantages and disadvantages to both. I like to use an omni in a good room with a choir, for example. However the cardioid M element is more common in most setups. http://en.wikipedia.org/wiki/Microphone_practice

It is interesting to note, however, that in most modern lower-cost (and even some insanely expensive) "MS" mics, the "MS" function is actually implemented with three cardioid mic elements. One facing left, one facing "mid", and the last facing right.  By adding L+C, you get a left polar pattern rather half-way between the front-facing capsule and the left facing capsule, and likewise on the right side.

The reason for using three cardioids vs. one cardioid and one figure-8 is that there are few decent studio-quality true figure-8 mic capsules with which to build microphones. I don't believe that Rode makes one, for example.

Rode have some multi-pattern mics, such as the K2 and the NT-2A. These can be switched to fig-8 pattern, so I expect that there are two diaphragm elements being selectively combined. Doug might be able to enlighten us on that one.

It would be interesting to discover how much difference you can hear between a LMR-style "MS" microphone and an ordinary "X-Y" coincident pair.

Tim Nielsen has an article in which he used Schoepps rigs for both MS and XY and claimed not to hear the difference http://designingsound.org/2011/08/tim-nielsen-special-ms-recording/

I have a mic rig that I've occasionally used for ensembles that comprises of a MS set in the middle, ORTF pair and spaced omni's on the same bar. I like to use the AKG C414XLII for MS, although I have used others as well, such as the AKG P420. The spaced omnis (about 70cm) are sometimes preferrable when you don't necessarily want a wide image, but want the extended LF response that omni's give you.

[DrGeoff]

I was testing out a new field recorder (Zoom H6) using its included MS microphone on the Sydney Wind Symphony at a concert last year and here you can hear the effect of an MS recording and the accurate position of each of the instruments in the stereo field.

The performance of this recorder is simply amazing. It seems to me that the highs are way softer than on H4, did you eq the track or do the H6 mics sound that way?

There is always post production
The raw files from each of the mics were combined in the usual MS mix with 3 channel strips. The resulting stereo group may have been slightly eq'd and level adjusted before printing, I'd have to go and check the mix notes to be sure.
The other recordings prior to this concert were all done using the R09HR recorder, which is a lot smaller and simply has two omni mics. The interesting thing to note is that, whilst the image is definitely a lot narrower, the distance between the front and back lines of the orchestra (ie clarinets to percussion) also appear to be smaller.

[Richard Crowley]

The MS mid mic can be either omni or cardioid. There are several articles about this. It depends on your application as to whether you want a rear rejection in the mono element (cardioid) or you want sound from the rear in the mono element (omni). The maths for the sides is the same. There are advantages and disadvantages to both. I like to use an omni in a good room with a choir, for example. However the cardioid M element is more common in most setups. http://en.wikipedia.org/wiki/Microphone_practice

we could call an MS with omni "center" something like "Super-8"  (not unlike hyper-cardioid). 

Every appearance is the the Rode models with "steerable" directionality are using the classic Heumann dual-capsule design with switchable/variable bias voltage.

[edy]

Hi folks,

Inspired by the excellent video by Dave and Doug, I put together this quick EXCEL Spreadsheet (attached) to allow you to see OMNI and FIGURE-8 microphones and also the MIX.

You can alter the mix ratio... Just set the first one (where it says 0.5) and it will automatically adjust the other to balance, so they both add up to 1 (or 100%). Watch the graph change!!!!!

Let me know how you like it...

OOPS... Just realized something, got my PHASE wrong on the reverse side of the mic... will have to alter my EXCEL SPREADSHEET... Hang on! :-)

[edy]

Hi folks,

I modified the Excel spreadsheet... having some issues but it looks closer now. The secret was my Cardioid calculation was wrong. I had to get a polar vector magnitude on the Omni and Figure-8 curves, add them up and reverse the sign during the negative (180-360) part of the trace. Hopefully it is closer now....

Feel free to modify the spreadsheet and improve calculations.

[Hideki]

And for everyone without Excel, you can do it on the web: http://fooplot.com/plot/ekis0gk6y1

[edy]

Hey Hideki,

Nice one... will have to check out Fooplot more. I still can't seem to get the extra "lobes" Doug was talking about in the video that go off in different directions on the negative-phase side. Maybe the formulas I used are too simple. Excel is harder to do polar plots, the Fooplot is much easier.

Maybe we can generate a cardioid microphone whose characteristics will appear like a Mandelbrot.  :-)